A methodology and a case-study for Network-on-Chipbased MP-SoC architectures

The many-core design paradigm requires flexible and modular hardware and software components to provide the required scalability of next-generation on-chip multiprocessor architectures. A multidisciplinary approach is necessary to consider all the interactions between the different components of the design. In this work a complete design methodology is proposed, tackling at once the aspects of hardware architecture, programming model and design automation. The proposed design flow has been used in the implementation of a multiprocessor Network-on-Chip based system, the NoCRay graphic accelerator. The system uses 8 Tensilica LX processors and has been physically implemented on a Xilinx Virtex-4 LX-160 FPGA reporting a 17.3M equivalent gate-count. Performance are compared with a commercial general purpose processor and show good results considering the low frequency of the prototype